Oecologia (2002) 131:20–26 DOI 10.1007/s00442-001-0846-8 Abstract Many arthropods produce clusters of eggs, but an unambiguous explanation for the evolution of egg clustering is still lacking. We test several hypotheses for the production of egg clusters by the predatory mite Iphiseius degenerans. This predator feeds on pollen, thrips larvae and nectar in flowers, but oviposits in clus- ters in tufts of leaf hairs (acarodomatia), where eggs run a lower risk of being killed by thrips, the prey of this predatory mite. The observed clustering is not caused by a shortage of oviposition sites; females preferably ovi- posit in a domatium containing eggs rather than in an empty domatium. To explain this preference, we first ex- amined the effect of egg clusters on the risk of cannibal- ism. We found that eggs are invulnerable to cannibalism, whereas larvae emerging from single eggs or from clus- ters were equally vulnerable. Subsequently, we consid- ered the killing of eggs resulting from counter-attacks by prey, i.e. the western flower thrips. We found no indica- tion that a cluster of eggs protects eggs from predation by thrips. However, when eggs were clustered in a dom- atium rather than scattered over domatia, the proportion of eggs killed by thrips was lower. Hence, oviposition in clusters has no effect on its own and oviposition in domatia reduces predation risk by thrips, but oviposition in clusters in domatia leads to a synergistic effect on the survival of predator eggs. This synergism probably arises because eggs in clusters within tufts of leaf hairs are more difficult for thrips to reach. These experiments highlight a novel explanation of egg clustering, i.e. adap- tation to counter-attacking prey. Moreover, they show that plant domatia protect predator eggs from predation. Keywords Domatia · Egg clustering · Predation · Counter-attacking prey · Cannibalism Introduction Oviposition in clusters is a common phenomenon in var- ious insect taxa (Parker 1965; Weseloh 1972; Stamp 1980; Ito et al. 1982; Chew and Robbins 1984; Damman and Cappuccino 1991; Stewart et al. 1991; McCauley 1992; Agarwala and Dixon 1993; le Masurier 1994; Codella and Raffa 1995; Hunter 1995; Clark and Faeth 1997, 1998). Several terms have been used to describe this form of oviposition, such as egg clumping, egg pool- ing, aggregation of eggs or batches of eggs, but defini- tions regarding these terms are absent or ambiguous. In this paper, an egg cluster is defined as a group of eggs, produced by one or more females, in which each egg touches the rest of the group through contact with at least one other egg. There are three major groups of explanations for the evolution of egg clustering. Some of these explanations are poorly supported by published data, while others are likely to be the consequence rather than the cause of such oviposition behaviour (Courtney 1984). The first group centres around the direct effects on the ovipositing female; by reducing search for oviposition sites and lay- ing many eggs together, the female saves energy, incurs a lower mortality risk during searching (by predation or adverse weather conditions), and/or reduces the risk of failing to find another oviposition site (Stamp 1980). These explanations implicitly assume that moving from one oviposition site to another has associated costs and that selection therefore acts to minimise such movement. The second group of explanations focuses on the ben- efits of clustering through effects on fitness of the larvae that hatch from the eggs. Larval survivorship and perfor- mance are thought to be enhanced by aggregation be- cause of predator saturation (Lawrence 1990), communal defence against predators and parasitoids (Labine 1968; Tostowaryk 1972; Stamp 1980; Chew and Robins 1984), an enhanced effect of warning coloration (Labine 1968; Stamp 1980), increased efficiency of thermo-regulation (Seymour 1974; Porter 1982), or increased efficiency of resource use (Young and Moffett 1979; Tsubaki and F. Faraji · A. Janssen ( ✉ ) · M.W. Sabelis Section Population Biology, Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, P.O. Box 94084, 1090 GB, Amsterdam, The Netherlands e-mail: janssen@science.uva.nl Fax: +31-20-5257754 POPULATION ECOLOGY Farid Faraji · Arne Janssen · Maurice W. Sabelis The benefits of clustering eggs: the role of egg predation and larval cannibalism in a predatory mite Received: 20 August 2001 / Accepted: 22 October 2001 / Published online: 30 November 2001 © Springer-Verlag 2001